Tuberculosis is a major killer of Human Immunodeficiency Virus (HIV)+ persons worldwide. Compared to the 10% lifetime risk of a PPD+ person developing tuberculosis, an HIV+PPD+ person has a 10% annual risk of this disease. The interaction between HIV and Mycobacterium tuberculosis is not well-understood. By necessity, many of the studies to date have been performed in vitro or in a natural infection human setting, due to lack of an appropriate animal model. The non-human primate model can be used to address this interaction using Simian (human) Immunodeficiency Virus (SIV/SHIV) and M. tuberculosis co-infections. Specifically, in this application we will address the serious problem of latent tuberculosis and mechanisms by which reactivation of latent tuberculosis can occur. Many people infected with HIV are already latently infected with M. tuberculosis, and reactivation can occur at any level of immunocompromise. Using a cynomolgus macaque model of low-dose M. tuberculosis infection recently developed in our laboratory, we will explore latent and reactivation tuberculosis. Our non-human primate model appears to mimic human latent tuberculosis. We will examine and compare reactivation of latent tuberculosis in the macaque model using three different immunocompromising strategies. The reactivation triggers we have chosen are CD4 T cell depletion by antibody, SHIV co-infection, and TNF-a neutralization. By comparing the effects of each strategy on latent tuberculosis, in terms of clinical, immunologic and pathologic parameters, we can gain an understanding of latent tuberculosis, and this knowledge will be useful in devising strategies to prevent reactivation. We will also learn about the mechanisms by which HIV leads to reactivation, by comparing each of our three models. These studies are the first to study reactivation in an immunologically tractable animal model that is similar to human latent tuberculosis.